Enterprise Storage Forum content and product recommendations are editorially independent. We may make money when you click on links to our partners. Learn More.
There are many types of computer memory, an umbrella term for data storage technology built into or externally connected to a computer device or server. Some are designed to be fast—the central processing unit (CPU) can access data stored there quickly—while others are designed to be low cost, so large amounts of data can be stored economically.
All computer memory types fall into one of two categories, primary or secondary, depending upon their purpose and use case. Knowing their different applications and strengths is essential to understanding how storage works and how to make the most of it.
Table of Contents
Primary Memory Types
Primary memory is the segment of computer memory that can be accessed directly for rapid processing and for quicker booting. It’s usually placed in close physical proximity to the CPU in order to minimize communication times. Examples of primary memory types include random access memory (RAM) and read only memory (ROM).
Both RAM and ROM are used to store data for direct access by the CPU. RAM is volatile, meaning its data is deleted once the system has been powered down. ROM is non-volatile, meaning it keeps its data after the system has been powered down.
| RAM | ROM | |
|---|---|---|
| Stands for | Random-access memory | Read-only memory |
| Storage volatility | Volatile | Non-volatile |
| Storage capacity | Up to 256GB per chip | up to 8GB per chip |
| Speed | Faster | Slower |
| Use | In standard computing operations | During the boot process |
| Types | DRAM, SRAM | PROM, EPROM, EEPROM |
| Cost | Higher | Lower |
| Chip size | Larger | Smaller |
Random Access Memory
Random-access memory (RAM) gets its name from the way the CPU accesses it and scans its lines randomly for the appropriate information rather than following strict directions. This is to equalize the time of access between all stored bits of data.
One of the defining features of RAM memory is that it’s incredibly fast—only marginally slower than the CPU itself. RAM is used to hold the data the CPU needs while it runs a program, which saves it time to reach the secondary storage every time it needs a piece of data or instruction.
However, RAM is also one of the most expensive components of a computer device, with cost usually measured per gigabyte. The high cost of RAM is the main reason why computers have to rely on both primary and secondary memory. There are two main types of RAM.
Dynamic RAM (DRAM)
The most common variety of RAM contains both a transistor and a capacitor in each of its cells. All computers must contain both RAM and DRAM to function and operate effectively. The oldest version of DRAM is known as single data rate (SDR) DRAM, but more recent versions like DDR2, DDR3, DDR4, and DDR5 tend to be more energy-efficient and more productive. The different types aren’t compatible, and computers can only use one type of DRAM.
Static RAM (SRAM)
SRAM is faster than either RAM or DRAM, but it’s considerably more expensive. It’s also bulkier, with six transistors in each cell. As such, SRAM is only used in small quantities, and almost always as cache memory to mediate the communication between the device’s main RAM and CPU and bridge the gap in clocking speeds. The fastest and smallest implementation of SRAM is as a cache memory inside the CPU’s core, working at the same speed to eliminate delays.
Read-Only Memory
ROM gets its name because data stored on it is immutable—it cannot be removed or changed using ordinary methods. Like RAM, it’s an incredibly fast type of computer memory that can be found in close proximity to the device’s CPU. But unlike RAM, ROM is non-volatile—it doesn’t lose its data when power is cut off—which it has in common with secondary memory.
When a device is turned on, or booted, the first thing the CPU reads is Instructions on the ROM, which usually contains “bootstrap code” that allows the computer to carry out the necessary steps for launching the operating system on the secondary memory.
There are three main types of ROM that can be found in all types of electronic devices, from game consoles to car radios and navigation systems.
Programmable ROM (PROM)
PROM is different from true ROM in that it is manufactured in its empty state instead of sold and distributed with instructions already programmed on it. PROM can be programmed after the fact using a PROM programmer or burner.
Erasable Programmable ROM (EPROM)
The information stored on EPROM can be erased and then reprogrammed. This can only be done by taking it out of the device and exposing it to ultraviolet light to erase it before burning new data to it.
Electronically Erased Programmable ROM (EEPROM)
EEPROM differs from both PROM and EPROM by not needing to be taken out to erase and reprogram. While the reformatting process can be slow, it’s not done often and usually only to update critical code such as firmware or BIOS.
Secondary Memory Types
Secondary memory is the persistent, non-volatile segment of computer memory not directly accessed by the CPU. It’s designed for more affordable long-term storage of large amounts of data. Examples include hard disk drives (HDDs), solid state drives (SSDs), cloud storage, and tape drives, to name a few.
Secondary memory tends to be more affordable than primary memory, and is available with much greater capacities. The different types of secondary memory enable users to store personal data and information along with software, applications, and services that can be accessed indirectly by the computer’s CPU through its RAM.
Hard Disk Drives (HDDs)
Hard disk drives are electro-mechanical data storage devices capable of keeping and restoring data over a long period of time. They’re built from a stack of rotating disks held in place by spindles. HDDs are the cheapest option per gigabyte compared to other long-term secondary storage, but they’re less durable.
HDDs have many moving parts, all of which have their own possible points of error. They’re also noisy, consume a lot of power, and can be slower to access and write compared to the speed of the average RAM. That said, HDDs are great for long-term storage and archiving. They’re often found in older and more budget-friendly desktop computers and laptops.
Solid State Drives
Solid-state drives are a type of semiconductor-based long-term, non-volatile storage. They use NAND flash memory in persistent data storage and are typically used as secondary storage in a computer to hold personal files. SSDs use flash memory to keep data in an integrated circuit. Data on an SSD can be written, read, duplicated, and transferred electronically without making noise or requiring movement.
Without the mechanical parts found in their HDD counterparts, SSDs are capable of operating silently and efficiently with minimal chances of physical damage. While they aren’t as durable, they can last up to five years, which means they aren’t as effective for archival and long-term storage. They’re faster and more compact than HDDs, but considerably more expensive and tend to be smaller in capacity.
Optical (CD or DVD) Drives
Optical storage is an umbrella term for storage that uses a laser to store and retrieve data from optical media such as CDs, DVDs, and Blu-ray discs. Most optical storage discs are read/write, allowing you to repeatedly erase and reburn new information onto their surface, but require a specialized drive.
An optical drive works by shooting a low-energy laser beam to scan the surface of the spinning optical storage media. How fast the writing and reading process depends on the spinning speed of the disc. On a micro level, optical discs have grooves and protrusions the optical drive can detect by analyzing how the laser reflects off of its surface.
Tape Drives
Tape drives are long-term, persistent, non-volatile data storage devices that use magnetic tapes to store, read, and write bits of data. They consist of an outer covering that protects a loop of flexible material that carries the data. Similarly to optical storage, tape storage cannot be accessed without specialized equipment. Tape drives use linear tape technology or a helical scan to read the magnetic markings on the tape.
Tape drives are ideal for long-term and archival storage, as the average unit can last for up to 30 years. They’re also cost-efficient and easy to maintain, making them suitable for keeping large volumes of data. That said, they’re bulky and incredibly slow—the drive has to rewind and go through the length of the entire tape to reach a specific point and pull data from it, unlike HDDs and SSDs that can be accessed more efficiently.
Storage Arrays
Storage arrays, also known as disk arrays, consist of a collection of fast-spinning HDDs, SSDs, or a hybrid of both storage types and are mostly used for storing object-based, file-based, or block-based data. They’re the most common alternative to server storage, as arrays combine multiple drives into a single, large-scale data storage system with centralized management.
Capable of storing and reliably maintaining petabytes of data at a time, storage arrays make data accessible by local client nodes as well as remote computers through an application programming interface (API) or dedicated graphical user interface (GUI) control panel. Storage arrays are employed by organizations, businesses, and enterprises with a large volume of data that needs to be managed centrally but accessed remotely by multiple users.
Network Attached Storage
Network attached storage, or NAS, is file-based storage connected to a wider computer network to give access to a wide range of users. While NAS is mostly used by large corporations and enterprises, it can be adapted to suit the needs of a small team of professionals. It’s ideal for situations where more than one device needs to access the same data simultaneously. It’s also great for collaborative work, as it only offers the latest version of a file.
Cloud Storage
In cloud storage, data is kept remotely on servers that can be accessed through an internet connection from anywhere, using any device. It’s widely used by both businesses and individuals to store data and files they need to access regularly, without having to carry them around locally or on external storage.
What Is the Difference Between Primary and Secondary Memory?
If the computer’s CPU directly communicates with a type of data storage, it’s considered primary storage and essential to the functioning and operation of the device. Secondary memory, on the other hand, encompasses a variety of storage device types not accessed directly by the CPU but rather through the primary storage. Here are a few other main differences:
- Volatility—RAM is considered volatile storage, where all data is erased if the hardware loses power, while secondary memory is capable of keeping its data even after a power loss.
- Speed—Primary storage needs to be fast in order to keep up with the rate of the CPU core, while secondary memory is relatively slower and not suitable for direct CPU access.
- Cost—Because primary memory storage is designed to be incredibly fast, it’s also expensive and sold with smaller storage capacities, while a single secondary storage device can reach terabytes in size and still be reasonably priced.
Differences Between Volatile and Non-Volatile Memory
Volatile memory requires power to maintain stored information. When a computer is switched off, or if there’s a power outage, any data stored in volatile memory is lost. It’s typically faster than non-volatile memory, it’s where data that needs to be accessed quickly—such as the operating system, running applications, and the data those applications use—is stored.
The main characteristic of volatile memory is its temporary nature. It’s used for data that needs to be accessed instantly but doesn’t need to be kept. Volatile memory also offers high-speed data access, making it ideal for tasks that require rapid data processing.
Non-volatile memory can retain stored information even when not powered. A primary example of non-volatile memory is the hard disk drive in a computer where data that doesn’t need to be accessed quickly—such as files, operating system, applications, and software—is stored. Non-volatile memory is characterized by its ability to retain data even in the absence of power. However, it’s typically slower than volatile memory, as it’s not designed for rapid data access.
Bottom Line: Understanding the Types of Computer Memory
Technology relies upon many types of computer memory for both short- and long-term storage. Some are designed for rapid access, while others are designed to more economically retain larger amounts of data for longer periods. Understanding the differences among the various types, as well as their applications and use cases, can help you better maximize your data storage and data management efforts.
Read Five Types of Enterprise Data Storage to learn more about the different ways businesses store and manage information.
